CONTRIBUTION TO THE VIEW HELD ON MICROHARDNESS OF ALKALI HALIDES

CONTRIBUTION TO THE VIEW HELD ON MICROHARDNESS OF ALKALI HALIDES

By

J.

J.

Department of Experimental Physics, Technical "Cniversity, Budapest (Recei.-ed December 15, 1971)

Presented by Prof. Dr. J. }LhRAI-ZDIPLE:i

Methods of hardness measurements haye much improyed in the recent decades. Progress is mainly in the more exact processes of measurements and in the detection of testing errors. A far less success was achieved by relating the yalues of hardness obtained by the different methods. This is not surpris- ing if we consider that the hardness number obtained by the difierent methods

of measurements (e.g. wearing out, scratching, different types of micro-inden- tation, etc.) are not necessarily related to the same fundamental physical properties of the material. The material does not deform to the same degree in the different methods of measurement and so it has a different measure of hardening. In spite of all these difficulties, hardness tests are widely imple- mented because of their many adyantages. Establishment of a relationship between the principal physical properties and the hardness number of the material would be of much help.

Many experimental results proye that the cation impurity increases the hardness of alkali halide crystals. The authors attribute the increased hard- ness to the interaction between the dislocation and the multiple point defects caused by doping [1, 2J.

The objective of the present work 'I-a;:; to inv(>stigate this question.

Experiments and results

The samples were cleft from :\"aCl single crystals eontammg different contaminations. Three groups of samples can be considered in view of con- tamination, such as: extrcmely pure (impurity is less than 10 -5 mol %):

purposefully doped (a kno,m quantity of cation impurity given to the extreme- ly pure basic material); and nominally pure ones. The single crystals in the first two groups were grown by Bridgman's method from a specially puri- fied ba;:;ie material.

*

* Also here we want to accentuate our thanks to R. Y OSZKA, senior lecturer to the Biophysical Institute at the University of }Iedical Sciences, Budapest. for placing at our disposal the samples of extreme purity as well as the ones purposefully doped with cation after the purification.

K Periodieu Polyteehnica El. XYI,~.

200 J. SARKOZI et at.

The nominally pure crystals were grown from a pro analysi basic material by the Kyropoulos method, and so they contained, besides the usual cation impurity mainly, anion and oxygen contaminations as well.

We cleaved uniform samples of 5 X 5 X 1 mm:^l size from crystds that were tempered beforehand in air at 650 QC. Heat treatment in nitrogen atmos- phere did cause no change of importance in the quantities me:::sured.

The point defect structure of samples with the given rate of impurity 'was changed by quenching from an elcvated temperature to room temperature [4] at such a speed that the concentration of the dislocation increased but slightly.

The Vickers microhardness value was chosen to charactprize the hard- ness of the quenched crystals. In order to have information for the interaction between point defects and dislocations we also examined the length of "rosetta"

appearing after a selective etching - around the indent er [5].

a) The Vickers microhardness test

Microhardness was measured at room temperature on samples prepared as indicated above by a Haneniann D32 microhardness tester that 'was set up on a Zeiss NEOPHOT microscope. Most tests applied a load of 4. p., and the microhardness vdue was calculated as an average of 40 indentations.

Loading was interpreted vs. diameter of impression and the rule by MOTT [6].

P = ad" was proved where P is the load, d is the indentation diameter, a and

1l are the structure sensitive coefficients. According to the investigations, both the a value and the microhardness yalue HJ.' obtained at a load of 4 p sho'w the same trends of changing as a function of quenching temperature, and this is why the micro hardness yalue obtained at an indentation of 4 p is used in the present study for ch:::racterizing the mechanical properties of crystals.

In Fig. 1 the microhardness as a function of quenching temperature is shown for four crystals differring by both impurity concentration and con- tamination.

Study of the cun:es has led to some unambiguous conclusions.

1. It can be ascertained that the Ca type diyalent cation impurity creatcs a maximum in the microhardness (Hr) YS. quenching temperature iTq) func- tion, at about the quenching temperature of 300 QC (curye 1 in Fig. 1).

2. The presence of TI as monovalent impurity in the crystal lattice makes a peak in the H,,- Tq curye at about 500 QC (Fig. 1 curye 2).

3. If the mono- and the diyalent cation impurities are present together in the (nominally pure) crystal lattice then both maxima develop, indicating that they exert their effect independently of the microhardness (Fig. 2).

4. For quenching temperatures over 600 QC, the microhardness increases independently of impurity contents (See Figs 1 and 2).

THE VIEW HELD O:V MICROHARD,YESS 201

H 34

(kp/mm2) 33 32 31 25 24 23 22 _£r 21

0 100 200 300 400 500 600 700 Tq [OC}

Fig. 1. Microhardness of quenched :i'\aCI samples as a function of the quenching temperature.

Impurity contents of the samples: 1 0.1 mol% CaCI_{2 ,} 2 - 0.1 mol% TICI, 3 extremely pure

H 25 (kp/mm2)

24 23 22 21

20 ^~ ______ ^~ __ ^~ ____ ^~ __ ^~ __ ^~ __ ^~ __ __

o

Fig. 2. The microhardness of a nominally pure quenched ::'IaCI sample as a function of quench- ing temperature

b) lYleasurement of "rosetta" length

SASKOLSKAYA and cO'workers demonstrated that the length of the so- called "dislocation-rosetta" might be characteristic of the hardness of alkali halide crystals. According to a later work [8] they have also measured the

"rosetta" length on quenched LiF samples containing lVIg contamination.

The measurements proved the close dependence of "rosetta" length upon the quenching temperature. We arrived at a similar conclusion for quenched NaCl samples with different rates of impurity. Tests were made according to [5].

8*

202 J. S.·{RKDZI ^ef at.

A typical series of test rcsults is seen in Fig. 3. The quenching temper&ture is plotted on the horizontal axis; as for the vertical one the relation of diagonal

"rosetta" length is depicted so that 'we gained on quenched and unqucnched samples of the same material. The "rosetto.s" were produced by a load of 4 p.

1/10

1,6

1,5

1,3

7,2

7,0 L-_"'--_"'--_~_-,--_ _ _ ~ _ _ _

o

Fig. 3. Relatiye rosetta lengths on quenched ::-IaCI samples as a fnnction of the quenching temperature. Impurity contents of the samples: 1 - 0.1 Il1olo~ CaCl o• 2 - nominally pure,

3 - extremely pure and 0.1 mol o ^fJ TICI

It is highly conspicuous at once that the "l'Osetta" lengths greatly change 'with the quenching temperature on samples 'with impurity of two valencies (curves 1 and 2 in Fig. 3). The rosetta lengths slightly react to tempering in samples made of an extremely pure basic material or a material doped 'with a monovalent cation (curve 3 in Fig. 3), and only a slow rise can be observed with increasing the quenching temperature. "Rosetta" lengths for samples contaminated 'with Ca + + show a very sharp change in growth at about a quenching temperature of 400 cC. Cun-es of a similar trend were obtained when "rosetta" lengths were measured at a constant indentation diamcter.

Discussion

The microhardness value, like other hardness characteristics presents the material's resistance to plastic deformation. Consequently, hardness is in close connection 'with the process of plastic deformation, thus, first and fore- most the plastic deformation has to be the point of scientific investigation.

To generate a plastic deformation - as it is 'well known - dislocations have to develop and propagate.

A great majority of research 'workers explain the change of hardness of the crystals by the change of energy demand for the dislocations to propa- gate. For example FLEISHER'S [2] interpretation for hardening upon diyalent

203 cation impurities, is that hardening is due to dislocations caught on complexes.

PRATT giv-es a similar explanation for his experimental results [9].

~ev-ertheless, our present results might permit to draw the conclusion that the fOH'going interpretation c:tu only be applied for hardness maxima formed at 300 °C, and that is unambiguously connected "with the diyalent cation impurity (Fig. 1 eun-e 1). The braking of disloc2tions caused by the

"complexes" appears by the small lengths of the "rosettas" and in the micro- hardness higher values. "When at a temperature over 300 cC the disintegration of complexes begins [10], the relative scale lengths of the "roscttas" increase (Curves 1 and 2 in Fig. 3), and the microhardness v-alue decreases.

The microhardness and the "rosetta" length curves' progress at higher quenching temperature ranges already points to the undecisiv-e importance of dislocation braking in the hardness yalues. This is proyed for example by the fact that no influence of monovalent cation impurity at the scale of the

"rosetta" lengths (that might be the characteristic for dislocation braking) can he observed (curye 3 in Fig. 3), while it can v-ery well be observed in the microhardness (cm'ye 2 in Fig. 1). This is also confirmed hy the phenomenon how the quenching - oyer 600 cc increases the crystals' micro hardness irrespectiH of their contamination (Fig. 1 and Fig. 2), while no intrinsic change was obseryed at the "rosetta" lengths scale.

From the foregoing ·we might draw the conclusion that in the hardness yalue another factor might play a role in addition to the braking of disloca- tions, i.e. the process of generation of dislocation concomitant of plastic defor- mation. Changes in the process of gener~ttion and multiplication of dislocations might much affect hardness values, and, at the same time, these changes do not necessarily appeal' in the length of "rosetta" [11].

There arc ohservations pointing out the importance of point def'~cts for the formation and multiplication mechanism of disloeations [1:2, 13]. In our case it is easily conceivahle that a tf'mpe'ring at ahout 500 cC create's a concli- tion for the lllonoyalel~t contamination \I-hicll has a hraking effect upon dis- loeation multiplication. Condensed yacancies might play the same role in crystals quenched from a temperature oyer 600:C.

A,aibhle data do not permit cv-ell a qualitatin: picture about the cor- relCition bet,n:eIl the generation of dislocations and the harcllle'ss value, thns, this is only to indicate the likelihood of phenomena.

SUllunarv

The microhardncss of quenched ::\aCl crystals extremcly pure, nominally pnre (con- taining both anion and cation impurities) and doped purposefully with monoyalcnt and diyalcnt cation, was inyestigated as a function of quenching temperature. It was obserycd tlll,t the microhardncss of crystals qaenched from a giycn temperature increased upon the effect of contamination present in the lattice, The explanation is connected with the genera- tion mechanism of dIslocations.

204 J. SARKOZI cl al.

References

1. A,lIIOaKOB, A. A.-Eyi1KO, B. M.-MaHYll,lOBa, t1. Jl.: B.lll51HlIe rrplIMecei1 H CTPYKTYP- HbIX ;:(e(jleKTOB Ha CBoi1cTBa HeMeTa.1JllP-leCKIlX KpHCTa.1J10B. <DpYH3e 1965. CTp. 26.

2. FLEISHER, R. L.: Acta :Met. 33, 3504 (1962).

3. VOSZKA, R.-TARJ.-ii'i, L-BERKEs, L.-KRAJSOYSZKY, J.: Kristall und Technik 1, 423 (1966 ).

4. KISS, J. L.-SARKOZI, J.-VA"i'iAY, L.: Phys. Stat. Sol. (b) 43, K157 (1971).

5. WacKo.lbCl{a51, M. n.-,UOOp}KaHCKlIlr, r. <D.: l{pHcTa~l'lorpaqJlI51, 7, 103 (1962).

6. MOTT, B. W.: Micro-indentation Hardness Testing, London. 1956.

7. WaCKOJlbCI,a51, M. n.: <DH3IlKa me,10".Hora,11I;:(HbIX l{pllcTaJl.10J3. PlIra 1962. CTp. 466.

8. EJ1IJrTaHOB, A. A.-ryceB, 9. E.-,Uoop/!{aHcKHJ:'r, r. <D.-CMYWKOB, 11. B.-TarneBa, 111. i\t-WaCKOJlbCI,a51, Nt n.: ,UHHa~ulKa ;:(llCJ10KaUIli1. XapbKOB 1968. CTp. 470.

9. PRATT, P. L.: Proc. Brit. Ceram. Soc. 1, 177 (1964).

10. GRVi'iDIG, H.: Z. Ph}"s. 158, 577 (1960).

11. GIL2IIAi'i, J. J.: J. Appl. Phys. 33, 2703 (1962).

12. npe;(BO;:(IITeneB, A. A.-Po/!{aHcKlli1, B. H.-CTenaHoBa, B. M.- TYMaHoBa, H. A.:

l{plIcTaJ1Jlorpa(jl1l51, 9, 695 (1964).

13. Ha;(r~pHbIH, 9. l\\.-CTerraHoB, A. B.: <DTT, 5, 1006 (1963).

J

l

Andnls TOTH Budapest XI., Budafoki u.

Dr. J

8, Hungary